- Home
- Skills
- Testing & Debugging
- LVGL MCP — Let Your Agent See the UI It Writes
LVGL MCP — Let Your Agent See the UI It Writes
Headless LVGL v9 UI development with real-time PNG rendering and touch interaction simulation.
Free
Works with the AI Agents you already use
LVGL MCP — Let Your Agent See the UI It Writes
Example session with this skill installed
Create a 320x240 settings screen with a 'Wi-Fi' toggle and a brightness slider. Render it to make sure the labels are aligned left.
- Read your context and instructions
- Compiled the lvgl mcp
I have written the C code for your settings screen and rendered it to verify the layout. The labels are aligned to the left using a flex layout, and the controls are centered vertically within their respective rows.
LVGL Settings Screen
#include "lvgl.h"
void create_ui(lv_obj_t *parent)
{
lv_obj_set_flex_flow(parent, LV_FLEX_FLOW_COLUMN);
lv_obj_set_flex_align(parent, LV_FLEX_ALIGN_START, LV_FLEX_ALIGN_CENTER, LV_FLEX_ALIGN_CENTER);
lv_obj_set_style_pad_all(parent, 10, 0);
lv_obj_set_style_row_gap(parent, 15, 0);
/* Wi-Fi Row */
lv_obj_t *row1 = lv_obj_create(parent);
lv_obj_set_size(row1, lv_pct(100), LV_SIZE_CONTENT);
lv_obj_set_flex_flow(row1, LV_FLEX_FLOW_ROW);
lv_obj_set_flex_align(row1, LV_FLEX_ALIGN_SPACE_BETWEEN, LV_FLEX_ALIGN_CENTER, LV_FLEX_ALIGN_CENTER);
lv_obj_t *label1 = lv_label_create(row1);
lv_label_set_text(label1, "Wi-Fi");
lv_obj_t *sw = lv_switch_create(row1);
lv_obj_add_state(sw, LV_STATE_CHECKED);
/* Brightness Row */
lv_obj_t *row2 = lv_obj_create(parent);
lv_obj_set_size(row2, lv_pct(100), LV_SIZE_CONTENT);
lv_obj_set_flex_flow(row2, LV_FLEX_FLOW_COLUMN);
lv_obj_t *label2 = lv_label_create(row2);
lv_label_set_text(label2, "Brightness");
lv_obj_set_style_margin_bottom(label2, 5, 0);
lv_obj_t *slider = lv_slider_create(row2);
lv_obj_set_width(slider, lv_pct(100));
lv_slider_set_value(slider, 70, LV_ANIM_OFF);
}
Next steps
- Verify the alignment in the attached PNG render.
- Test the toggle interaction by calling
simulate_lvgl_uiwith atapat the switch coordinates. - Copy the
create_uifunction into yourmain.cor UI task.
Connects securely to your tools. The creator never sees your data.
About this skill
(Headless Render + Touch Simulation)
Your agent writes LVGL v9 create_ui() C code; lvgl-mcp compiles and renders it headlessly to a PNG, so the agent can actually look at the screen, fix what's off, and iterate — no display, no device, no flashing. A touch simulator replays taps, swipes and drags and returns screenshot sequences to verify interactions. Runs 100% locally: a stdio MCP server driving your own Docker daemon. Works with Claude Code, Cursor, Copilot and other MCP clients.
What's included
- The lvgl-mcp repository (https://github.com/airpocket-soundman/lvgl-mcp):
mcp_server/— FastMCP stdio server with three tools:render_lvgl_ui,simulate_lvgl_ui,lvgl_widget_cheatsheet.renderer/— Docker image (Debian + LVGL v9.4.0 prebuilt) with the headless renderer and touch-simulator harnesses; only yourui.cis recompiled per run, so iterations take seconds.setup.sh+.mcp.json.example— one-command setup and client wiring.examples/— a hello-world UI and simulator screenshot sequences.
- The SKILL.md, which teaches the buyer's agent when and how to drive the server (setup, the
create_ui()contract, the touch-action DSL, responsive-layout guidance). - Any canvas size (320×240 default — small-panel size — up to large panels such as 1280×720), Material Icons font included, gcc errors returned as readable text for the agent's fix loop.
Requirements
- Docker (local daemon), Python 3.10+, an MCP client.
- v1 scope: rendering + touch simulation only. Flashing real hardware is out of scope; the validated
create_ui()code drops into the buyer's own firmware project unchanged. - Needs access to: the lvgl-mcp MCP server, installed locally by the buyer from the public GitHub repo above.
Pricing note: this skill is free for now while I gauge interest. If there is real demand, I plan to switch it to a paid listing later — early users, thanks for trying it.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Testing & Debugging.
Requires Docker daemon and local Python environment. Supports LVGL v9 only. No support for hardware flashing or device-specific drivers.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 23 days ago
- Free with an Agensi account
- Accounts
GitHub
- Software
- LVGL MCP
Python
Git
Docker
Docker must run as a local daemon.
Permissions
Allowed Hosts
Needs access to: the user's own lvgl-mcp MCP server installed locally from https://github.com/airpocket-soundman/lvgl-mcp, requiring a local Docker daemon.